Sierra-1 NrT measurements and placement observations (off-axis). Maybe discovery?

[GirgleMirt]

Ok, so I've done a few measurements on the Sierra-1 and confirmed a few things: The treble is slightly 'hot' about 10khz (as per Ascend measurements in this thread (http://forum.ascendacoustics.com/showthread.php?t=5073)) and for some reason, as with it seems all 3 other speakes I have on hand, around 300hz to 2.25khz seems a bit lower... maybe it's my mic or measurement artifacts (cancellation from floor?), but anyway, after a bit of placement trial and error, the Sierras ended with very slight toe in, which reduces a little the high end and seems to give the more balanced results.

I had noticed a trend with speaker designs to have speakers with the front baffle in angle, for example, if the speaker points left, a regular baffle would be: | and the angle I'm talking about would be like / where the tweeter is further from the listener than the woofer, tweeter pointing upwards (but less angled). So last night I thought why not try it! So I simply angled up the speakers by putting foam under the front, toed in a little bit more towards me to compensate, listened maybe 5 minutes and it didn't sound bad at all! :eek:

So I did a few measurements, and with my really modest measuring gear, the measurements really don't look bad at all, seemed to even the FR a little, looks even better than straight on or even slightly better than off axis sideways! Which is really puzzling... (floor reflections reduced from angle?) So I'm wondering whether it's just due to measurements artifacts, or if this could actually be a good way to position the Sierra NrTs? Anyhow, I guess I'll try it for a bit and try to see if it's really a good idea, but this seems promising! :cool: Don't people angle up baffles to help with the time coherency?

Thoughts? Anyone else feels like exploring this? :)

I think it might not be a good idea to post in room measurements because of poor gear and so 'flawed' measurements which might just give false ideas? They're just pink noise I think it was, Behringer's DEQ2496's RTA & ECM8000 measurement mic, which I'm noticing now looking at the model number seems to add a slight (1dB) 'boost' from 3-20khz...

[GirgleMirt]

ok to get the ball rolling here are the measurements. I didn't write the angle because I didn't measure it exactly, speakers were set on stands in center of room with mic at about 1 meter (pink noise averaged):

up; speakers angled up so tweeter points bit upwards, maybe 10 & 20 degrees. Side; off-axis, maybe 15-25-35 I think.. Or somewhere along there..

As I said, the up two looks like it has the flattest overall FR with the treble in relation to the mids. If you take the ECM8000 bump (http://www.michnicki.com/diy/ecm8000-Manuf.FR.jpg) into account, hard to say but I think it would be pretty flat... Maybe someone with better measuring equipment could take a jab at it? ;) :D

But looking at my measurements, front '0' there's an upper 'plateau' starting at about 2.5k, and on average, 315hz to 2.5k seems to be on a lower plateau on average... up 2 seems to take these two plateaus and merge them to one, from about 315 to 5k, all the while reducing the upper treble tilt. If you compare to side 2 and side 3, the FR of up 2 seems better...

Oh and btw, these are the NrT Sierras.

[GirgleMirt]

Hmmm... One thing I didn't measure was off axis while tilted upwards... Hmmm... Feels like I'm talking to myself...

[GirgleMirt]

Did a bit of A/B comparison before supper, and actually, pointing upwards at about 15 degrees seems to be my new favorite placement! :eek:

It looks a bit weird with the foam to tilt, but for some reason it seems to reinforce bass (give a more surrounding sound?), seems to also bring the treble/mids more into 'focus' (on the same vertical plane or depth; bit less front/back depending on FR). I've compared vs with some toe in & basically no toe in. Basically no toe in imho sounds better than toed in, bigger soundstage and 'smoother' highs & mids, also bass bit stronger. Pointing upwards with some toe slight toe in seems to even and somewhat increase SS which is weird, and like I said, I think seems to improve highs/mids; bit somehow smoother and less 'depth' variation depending on the notes?

It's almost like rediscovering the NrTs!!! :cool: Quite weird.. I guess more listening is in order. :)

[hearing specialist]

Ok, so how did front 0 up 1 sound (first response graph) vs. the very last RTA response graph? The very first response had some serious 20K action going on, in fact all the higher freqs are stretchin' their legs there. Same disc, same song maybe use Zac Brown The Foundation intro to "Free".

[GirgleMirt]

Ok, so how did front 0 up 1 sound (first response graph) vs. the very last RTA response graph?
Like noise... Well they didn't really sound like anything, because that was basically a single speaker sitting in the center of the room playing pink noise at a mic located 1m away. :p

But "Front 0" means the speakers would be directly pointed at the listener, which just isn't a good idea in case of NrT I think, well in my room at least, maybe depends on that too... With toe in about 1/2 there (between straight ahead and pointing at listener), so that speakers point slightly outside listener, which corresponds somewhere along Front1/2, probably closer to "Front 1", it definitely sounds not bad but I can sometimes hear the 'high end energy' (mids/treble) and I prefer almost straight ahead which sounds a bit 'milder' or 'smoother'. "Front 1", the high end will sound well crisp, but it can sound a bit 'too much' sometimes (for instance, Pink Floyd tracks below). It's generally ok, but overall I think I prefer almost straight ahead which 'relaxes' things a little.


The very first response had some serious 20K action going on, in fact all the higher freqs are stretchin' their legs there.
Yep... As I mentioned, the microphone used probably has something to do with that. (ECM8000 link (http://www.diyaudio.com/forums/multi-way/96133-measuring-harmonic-distortion-ecm8000-3.html)):
http://i44.photobucket.com/albums/f44/bidland/Relativsammenlikning.gif

Maybe reflections too... Or could be a number of other artifacts... Like I mentioned, I measured 3 other set of speakers and they were all somewhat lower between 300 and 2000hz... Like I said, my measurements should really not be used as reference or anything, they're made using very modest gear, in room, just playing pink noise and the FR being averaged... Really not reference measurements... Just for the sake of comparing off axis horizontal vs vertical and try to figure out of angling the NrTs upwards make any type of sense...


Same disc, same song maybe use Zac Brown The Foundation intro to "Free".
Well this wouldn't work for measurements, but yeah real music is the reference when listening to the speakers and not measuring them.. hehehe

Today:
Dire Straits: Sultans of Swing
king crimson: thrak (One time (http://www.youtube.com/watch?v=RhJKCW2_w3k&feature=related) & walking on air (http://www.youtube.com/watch?v=2DdIUWus1gY&feature=related))
Pink Floyd: Great gig in the sky, time, One slip, Money intro.
Kawai Kenji: Ex: Unmei no Yoru (http://www.youtube.com/watch?v=QBBBGGvrz0Y)
Yoko Kanno:Dance of Curse (http://www.youtube.com/watch?v=O_49H_N43jQ)

Also have Chesky & other 'reference' quality recordings but I think these worked fairly well to hear the differences between the placement options.

Anyhow, I really don't pretend to have golden ears, and for instance I find speaker placement extremely challenging because moving them around changes a number of things and I have a very difficult time of trying to remember everything and comparing them ... But, like I said, my feelings were that orienting the Sierra NrTs towards the ceiling (10-15 degrees?) and toeing them slightly so that both speakers point 'outside' of the listener seemed to be most enjoyable today...

And on a side note, then I put on the 340SEs and boy these are really fantastic... These keep surprising me! 340SE measurements 1m directly in front for comparison (next to a side wall though...)

[Blutarsky]

I raised the front of my Sierra-1 speaker stands when I set them up. Subjectively, they sounded better. I have used this method on several pairs of speakers. I use museum putty to hold the speakers on.

Maybe it is the coffee table in front of the couch?

It is certainly an easy thing to experiment with.

B

[GirgleMirt]

Interesting! So yeah pointing upwards definitely alter the floor reflections, as well as the direct response from the tweeter and woofer...

I remember reading that angling a front baffle or recessing drivers do not make time coherent speakers. I think the crossover's effect also has to be taken into affect. But that said, is it possible that this (time coherency) also improves?

But looking at the graphs, I think it somewhat supports what my listening impressions... Around 2.5khz, the FR seems more even and definitely a bit flatter relative to the upper end... On the side measurements that region seem to be a bit like a top or start of a hill...

I would be very curious to see some Ascend measurements on this... Dave? Pretty please? :D

[curtis]

I remember reading that angling a front baffle or recessing drivers do not make time coherent speakers. I think the crossover's effect also has to be taken into affect. But that said, is it possible that this (time coherency) also improves?
I would think certainly changes the phase, but I doubt it improves since the desired phase is engineered into the crossover for the baffle to be vertical.

In essence, you are measuring the vertical off-axis response, correct? There are measurements posted for that.

[GirgleMirt]

I would think certainly changes the phase, but I doubt it improves since the desired phase is engineered into the crossover for the baffle to be vertical.

In essence, you are measuring the vertical off-axis response, correct? There are measurements posted for that.

For the NrT? The two set of measurements posted in the other thread where these ones:
http://forum.ascendacoustics.com/showthread.php?p=35068#post35068
http://forum.ascendacoustics.com/showthread.php?p=35500#post35500

Neither seem to include the same measurements as the regular sierra (http://www.ascendacoustics.com/pages/products/speakers/SRM1/srm1meas.html) and I couldn't find a link to the equivalent NrT measurements... Are similar measurements available anywhere?


In essence, you are measuring the vertical off-axis response, correct?
Essentially, yes and no. Sure, the vertical off-axis would be a start, but as mentioned, the horizontal off-axis response is also very important and when the speaker is tilted upwards, then the horizontal off-axis response also change... Is it then decent, good, bad? :confused:

That's why I mentioned that I didn't do the tilted + off axis measurements myself, and thinking back they would give a more complete picture of the viability of angling them back...


I would think certainly changes the phase, but I doubt it improves since the desired phase is engineered into the crossover for the baffle to be vertical.
Correct me if I'm wrong, but doesn't the Sierra-1 uses the OPPIX which takes into account the phase (or tries/optimizes it?), while the Sierra-1 NrT uses a 'regular' 2nd order crossover which inherently isn't perfectly phase coherent?

Searching:

The new tweeter allows us to do a few things differently in the crossover, including more precise control of the high-pass filter slope which allows for improved phase integration with the woofer (the stock Sierra-1 already has excellent phase integration but the NrT is even better).

This tweeter also allows us to use a different crossover topology with different slopes and not only is the difference in the highs dramatic -- the midrange is remarkable and the Sierra-1 NrT will be about 1/2dB more efficient with a frequency response that is now more linear than our CBM-170 (which is a remarkable accomplishment).

Hmmmm... Couldn't find details on crossover type... So yeah... I would imagine since most put the tweeter a bit backwards on supposedly time coherent designs, that although the NrT crossover is phase optimized, it probably isn't phase perfect... But the rear angle to make it phase perfect probably isn't a lot and might just turn out to be a few degrees... Anyhow, I'm not really preoccupied about it, I think I understand the basic of the principle, but the differences are so small imho in terms of ms that I don't think you could really perceive it... The fact that I can't give a straight/confident about which placement/position is best probably illustrates this...

Wow so I guess the original question a simple question with a very complex answer...

[GirgleMirt]

ok last bump ;) I highly suggest trying to angle up the Sierra NrTs! Some listening tonight again, and imho it sounds really good... I guess experimentation to find the perfect toe in & upwards tilt would be necessary for the 'perfect' position, but I'm fairly confident now that I prefer the sound (seems to sound better) pointing upwards... :cool:

Maybe I'll try to do more measurements with different tilts & off-axis combinations... At listening position too, see how all compare in practicality...

[hearing specialist]

Very interesting finds, i've been subscribed and reading pondering the thoughts of how, why, and what the heck???

Let me throw this at you, we have talked about spikes, we have talked about sound mats or benefits of decoupling so let me share a thought on something I experimented with. Stage set...burbur carpet on the thick side, spiked mains on said carpet my Advanced MCACC had a crazy 2K spike and a huge need for correction at 250hz as well. I place said mains with spikes on 2 2"x8"x12" mason pieces splitting the difference with each main so entire enclosure is sharing the 2 pieces each placing the spikes about 3/4's of the way on each outside mason piece butted up together. Now Advanced MCACC needs very little correction in any of the highs, my crazy 2K thing is gone and little EQ correction is needed and now 250hz also doesn't need very much EQ'n to correct. So now looking at the EQ screen shows less than 3db of correction at any available freqs with all setup the same and variables of change minimized. There is a sonic and measured difference especially within the high freqs that all sound needed less correction with this experiment. Go figure :D

[GirgleMirt]

I'd say "very cool! :D So I'm not crazy! ;)", but just to be clear, what did you do exactly?

I place said mains with spikes on 2 2"x8"x12" mason pieces splitting the difference with each main so entire enclosure is sharing the 2 pieces each placing the spikes about 3/4's of the way on each outside mason piece butted up together. Now Advanced MCACC needs very little correction in any of the highs, my crazy 2K thing is gone and

If I understand correctly, you just put the stands on granite slabs (or similar), and that was it? Or did you angle them up? :confused: Don't really understand that bit :( Or you put 1/2 the spikes on a slab to angle the speakers?

[hearing specialist]

Yep, they are concrete pieces painted black. Each tower has 2 each sitting still on spikes. The carpet with the spikes did give me an angle which I always thought was cool but with that type of thick carpet I always thought they tilted one way or the other and if I looked long enough I would start analyzing it way to much and measuring to see how much tilt or is it just my head playing games with my eyes. So, to rid the tilt in my case to ease my looking at it vs. watching tv I wanted to forever fix and especially after I saw the marks within the carpet they were leaving. Hockey pucks sounded horrible vs. spikes, no spikes with enclosures sitting flat on the carpet looked way worse and sounded bad as well and needed so much eq to smooth. So in reality the spikes on deep carpet did couple netting that aweful 2k and 250 hz. The clarity on these mason pieces was very obvious and since it needs less to correct makes one wonder which was correct to begin with. Logic got in my way maybe and my gut was telling to place on a surface that was never going to vibrate.

[GirgleMirt]

Hmmm! Ok I think I got it now! So basically, if you looked on top of one of your speakers [: :] (the brackets representing the towers looked at from the top, pointing to left lets say, dots representing the spikes), you took four pieces of granite, two for each speaker, and put the towers on top; So looking at the granite from the same angle: [ ][ ] and put the towers on top, with spikes. So you have the granite, no spikes, on carpet, and towers on top of granite: Somewhat like this: [[: ][ :]] And that improved things, and the angle of the speakers didn't change.

What are your mains btw? Sierra-1 NrTs or towers? If NrTs, Ascend stands filled?

I guess in my case, I have filled Ascend pedestal stands, on hardwood floor, and am using no spikes, simply some slim table mat that gets 'squished' between floor/stands. But a variable that gets added when angling the speakers back is the 1.5 inch foam placed in front to angle the speakers to angle them up... Maybe the foam itself has some impact...